Files
loc_az_hci/scripts/vm-management/monitor/monitor-and-complete.sh
defiQUG c39465c2bd
Some checks failed
Test / test (push) Has been cancelled
Initial commit: loc_az_hci (smom-dbis-138 excluded via .gitignore)
Co-authored-by: Cursor <cursoragent@cursor.com>
2026-02-08 09:04:46 -08:00

117 lines
2.8 KiB
Bash
Executable File

#!/bin/bash
source ~/.bashrc
# Monitor VMs and Automatically Complete Tasks When Ready
# This script continuously checks VM readiness and runs complete-all-vm-tasks.sh when ready
set -e
# Colors
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
RED='\033[0;31m'
BLUE='\033[0;34m'
CYAN='\033[0;36m'
NC='\033[0m'
log_info() {
echo -e "${GREEN}[INFO]${NC} $1"
}
log_warn() {
echo -e "${YELLOW}[WARN]${NC} $1"
}
log_error() {
echo -e "${RED}[ERROR]${NC} $1"
}
log_step() {
echo -e "${BLUE}[STEP]${NC} $1"
}
declare -A VMS=(
[100]="cloudflare-tunnel:192.168.1.60"
[101]="k3s-master:192.168.1.188"
[102]="git-server:192.168.1.121"
[103]="observability:192.168.1.82"
)
SSH_USER="${SSH_USER:-ubuntu}"
SSH_KEY="${SSH_KEY:-$HOME/.ssh/id_rsa}"
CHECK_INTERVAL=30 # Check every 30 seconds
MAX_WAIT=3600 # Maximum wait time: 1 hour
check_all_vms_ready() {
local all_ready=true
for vmid in 100 101 102 103; do
IFS=':' read -r name ip <<< "${VMS[$vmid]}"
# Check ping
if ! ping -c 1 -W 2 "$ip" > /dev/null 2>&1; then
all_ready=false
return 1
fi
# Check SSH
if ! ssh -o ConnectTimeout=3 -o StrictHostKeyChecking=no -i "$SSH_KEY" "${SSH_USER}@${ip}" "echo 'OK'" > /dev/null 2>&1; then
all_ready=false
return 1
fi
done
return 0
}
main() {
echo "========================================="
echo "VM Monitor - Auto-Complete Tasks"
echo "========================================="
echo ""
log_info "Monitoring VMs for readiness..."
log_info "Will automatically run tasks when all VMs are ready"
log_info "Checking every $CHECK_INTERVAL seconds"
log_info "Maximum wait: $MAX_WAIT seconds (1 hour)"
echo ""
if [ ! -f "$SSH_KEY" ]; then
log_error "SSH key not found: $SSH_KEY"
exit 1
fi
local start_time=$(date +%s)
local check_count=0
while true; do
check_count=$((check_count + 1))
local elapsed=$(($(date +%s) - start_time))
if [ $elapsed -gt $MAX_WAIT ]; then
log_error "Maximum wait time exceeded"
exit 1
fi
echo -n "[Check $check_count] $(date '+%H:%M:%S') - "
if check_all_vms_ready; then
echo ""
log_info "✓ All VMs are ready!"
echo ""
log_step "Running complete-all-vm-tasks.sh..."
echo ""
export SSH_KEY="$SSH_KEY"
export SSH_USER="$SSH_USER"
./scripts/complete-all-vm-tasks.sh
exit $?
else
echo "VMs not ready yet... (elapsed: ${elapsed}s)"
sleep $CHECK_INTERVAL
fi
done
}
main "$@"